diff options
| author |  Gael Guennebaud <g.gael@free.fr> | 2019-01-09 15:25:17 +0100 |
|---|---|---|
| committer | Gael Guennebaud <g.gael@free.fr> | 2019-01-09 15:25:17 +0100 |
| commit | e6b217b8ddf533de9bacc46aae2db6de78581056 (patch) | |
| tree | ac2ef320056bf2698ea021412198ff6609137a0a /Eigen/src/Core/arch/SSE | |
| parent | e70ffef9678f86ef465e93b89351e812ab47311d (diff) | |
bug #1652: implements a much more accurate version of vectorized sin/cos. This new version achieve same speed for SSE/AVX, and is slightly faster with FMA. Guarantees are as follows:
- no FMA: 1ULP up to 3pi, 2ULP up to sin(25966) and cos(18838), fallback to std::sin/cos for larger inputs
- FMA: 1ULP up to sin(117435.992) and cos(71476.0625), fallback to std::sin/cos for larger inputs
Diffstat (limited to 'Eigen/src/Core/arch/SSE')
| -rwxr-xr-x | Eigen/src/Core/arch/SSE/PacketMath.h | 11 |
1 files changed, 11 insertions, 0 deletions
diff --git a/Eigen/src/Core/arch/SSE/PacketMath.h b/Eigen/src/Core/arch/SSE/PacketMath.h index 3e7a75bc0..0003be43b 100755 --- a/Eigen/src/Core/arch/SSE/PacketMath.h +++ b/Eigen/src/Core/arch/SSE/PacketMath.h @@ -812,6 +812,17 @@ template<> EIGEN_STRONG_INLINE int predux_max<Packet4i>(const Packet4i& a) #endif // EIGEN_VECTORIZE_SSE4_1 } +// not needed yet +// template<> EIGEN_STRONG_INLINE bool predux_all(const Packet4f& x) +// { +// return _mm_movemask_ps(x) == 0xF; +// } + +template<> EIGEN_STRONG_INLINE bool predux_any(const Packet4f& x) +{ + return _mm_movemask_ps(x) != 0x0; +} + #if EIGEN_COMP_GNUC // template <> EIGEN_STRONG_INLINE Packet4f pmadd(const Packet4f& a, const Packet4f& b, const Packet4f& c) // { |